British Journal of Anaesthesia 113 (5): 85564 (2014)

Advance Access publication 30 June 2014 . doi:10.1093/bja/aeu202

PAIN

Perioperative gabapentin reduces 24 h opioid consumption

and improves in-hospital rehabilitation but not post-discharge
outcomes after total knee arthroplasty with peripheral
nerve block
H. A. Clarke1,2,4*, J. Katz1,4,5, C. J. L. McCartney2,4, P. Stratford6,7, D. Kennedy6, M. G. Page5, I. T. Awad2,4, J. Gollish3
and J. Kay8
1
Department of Anaesthesia and Pain Management, Toronto General Hospital, Toronto, ON, Canada
2
Department of Anaesthesia and 3 Department of Surgery, Sunnybrook Health Sciences Centre, Toronto, ON, Canada
4
Department of Anaesthesia, University of Toronto, Toronto, ON, Canada
5
Department of Psychology, York University, Toronto, ON, Canada
6
School of Rehabilitation Science, 7 Department of Clinical Epidemiology and Biostatistics and 8 Department of Anaesthesia, McMaster

Downloaded from http://bja.oxfordjournals.org/ by guest on October 14, 2016

University, Hamilton, ON, Canada
* Corresponding author. E-mail: hance.clarke@utoronto.ca, hance.clarke@uhn.ca

Editors key points
Gabapentin is known to
improve postoperative
pain scores and to reduce
opioid consumption after
surgery.
The authors studied the
effect of perioperative
gabapentin on knee
function after
arthroplasty.
They were unable to show
an improvement in knee
function or
movement-evoked pain.
Early analgesic
consumption and knee
range of motion
(secondary outcomes)
were improved.
Background. This study was designed to determine whether a 4 day perioperative regimen of
gabapentin added to celecoxib improves in-hospital rehabilitation and physical function on
postoperative day 4 and 6 weeks and 3 months after total knee arthroplasty (TKA).
Methods. After Research Ethics Board approval and informed consent, 212 patients were enrolled
in a randomized, double-blinded, placebo-controlled study. Two hours before surgery, patients
received celecoxib 400 mg p.o. and were randomly assigned to receive either gabapentin 600
mg or placebo p.o. Two hours later, patients received femoral, sciatic nerve blocks, and spinal
anaesthesia. After operation, patients received gabapentin 200 mg or placebo three times per
day (TID) for 4 days. All patients also received celecoxib 200 mg q12 h for 72 h and i.v. patient-
controlled analgesia for 24 h. Pain and function were assessed at baseline, during
hospitalization, on postoperative day 4 (POD4), and 6 weeks and 3 months after surgery.
Results. The gabapentin group used less morphine in the first 24 h after surgery [G38.3 (29.5
mg), P48.2 (29.4 mg)] (P,0.0125) and had increased knee range of motion compared with
the placebo group in-hospital (P,0.05). There were no differences between groups in favour of
the gabapentin group for pain or physical function on POD 4 [95% confidence interval (CI):
pain: 21.4, 0.5; function: 26.3, 2.0], 6 weeks (95% CI: pain: 0.1, 1.9; function: 20.2, 6.5) or 3
months (95% CI: pain: 20.2, 1.7; function: 22.2, 4.3) after TKA.
Conclusions. In the context of celecoxib, spinal anaesthesia, femoral and sciatic nerve blocks, a
dose of gabapentin 600 mg before operation followed by 4 days of gabapentin 200 mg TID
decreased postoperative analgesic requirements and improved knee range of motion after
TKA. Gabapentin provided no improvement in pain or physical function on POD4 and 6 weeks
or 3 months after surgery.
Keywords: functional outcomes; gabapentin; multimodal analgesia; pain; patient-reported
outcome measures; physiotherapy; total knee arthroplasty; TKA
Accepted for publication: 14 March 2014

Currently, over 500 000 total knee arthroplasties (TKAs) are
performed in North America annually.1 Osteoarthritis and de-
struction of the knee joint remain the primary indications for
total joint arthroplasty.2 Pain is often successfully treated
with non-steroidal anti-inflammatory drugs (NSAIDs) which
are beneficial early in the disease process; however, once
pain becomes severe with significantly compromised physical
function and significant radiographic evidence of joint space
destruction, patients are offered TKA.3
In recent years, gabapentin (a structural analogue of
g-aminobutyric acid), an anticonvulsant that binds to the
a2D subunit of voltage-dependent calcium channels in acti-
vated neurones, has been used widely as an adjunct for the
treatment of acute post-surgical pain. Meta-analyses have

& The Author 2014. Published by Oxford University Press on behalf of the British Journal of Anaesthesia. All rights reserved.
For Permissions, please email: journals.permissions@oup.com
BJA
confirmed the efficacy of gabapentin in reducing postoperative
opioid use and pain scores.4 6 Studies have examined various
regimens of co-analgesics after TKA7 8 but little is known about
the use of perioperative gabapentin for TKA on pain and func-
tional outcomes. A small, open label study found that patients
who continued to receive postoperative gabapentin 200 mg
three times per day (TID) for 4 days after operation used less
morphine and demonstrated improved early functional recov-
ery after TKA.9 In contrast, the addition of a 600 mg preopera-
tive dose of gabapentin followed by a 2 day regimen of
gabapentin 200 mg TID to a robust multimodal regimen did
not affect opioid consumption or pain scores after TKA.10
As one of the primary goals of TKA is to improve physical
function, this surgical model provides an opportunity to study
the effects of gabapentin on functional outcomes. Whereas
trials have demonstrated a reduction in post-surgical pain
with gabapentin,11 13 it remains to be seen whether this early
gain in pain translates into accelerated recovery in-hospital,
improved functional status after hospital discharge, or both.
Functional measures used herein include both patient-reported
outcome measures and performance-based measures of phys-
ical function.14 The Western Ontario and McMaster Universities
Osteoarthritis Index (WOMAC LK3.1)15 16 is a commonly used
patient-reported outcome measure used after TKA. Standar-
dized physical performance measures such as the timed up
and go (TUG) test,17 18 the timed stair test,19 20 and the six
minute walk test (6MWT)21 have been used to measure function
from baseline to 3 months after surgery and all have demon-
strated reliability and sensitivity to change within the total
joint arthroplasty population.22
Neuraxial anaesthetic techniques have become common
practice for lower limb arthroplasty.8 The addition of femoral
and sciatic nerve blocks for TKA has demonstrated a reduction
in postoperative opioid consumption and pain scores. Previous
work showed that a single shot femoral nerve block provided
superior analgesia compared with patient-controlled anal-
gesia (PCA) using morphine and pain relief equivalent to that
of continuous femoral block and epidural analgesia.23 The
present study evaluated the effect of gabapentin on pain and
recovery after TKA, within the context of multimodal analgesia
in which all patients received a preoperative spinal anaesthetic
(bupivacaine), non-steroidal anti-inflammatory medication
(celecoxib), and peripheral femoral and sciatic nerve blocks
with ropivacaine.
We do not know the optimal dosing and duration of peri-
operative gabapentin required to improve functional outcomes,
especially in the context of a clinically relevant perioperative an-
algesic regimen. A meta-analysis suggested that gabapentin
may prevent the development of chronic post-surgical pain;24
however, this hypothesis remains untested. The primary
purpose of this randomized, double-blinded, placebo-controlled
study was to examine whether in the context of preoperative
spinal anaesthesia, femoral and sciatic nerve blocks, and cele-
coxib co-administration, a 4 day perioperative regimen of gaba-
pentin vs placebo improves knee function on performance and
self-reported measures of physical function, and movement-
evoked pain on postoperative day 4 (POD4) and at 6 weeks
856
Clarke et al.
and 3 months after surgery. Secondary endpoints examined
whether this regimen improves in-hospital knee range of
motion (POD14) pain scores at rest and opioid consumption.
Methods
Patient sample and recruitment procedures
The study was approved by the Sunnybrook Health Science
Centre Research Ethics Board and all patients gave informed
written consent to participate. Patients between the ages of
18 and 75 years with an ASA physical status score of I, II, or
III undergoing TKA were eligible to participate. Patients were
not eligible if they had a known allergy to any of the medica-
tions being used, a history of drug or alcohol abuse, a history
of being on chronic pain medications (e.g. slow-release pre-
parations of opioids), rheumatoid arthritis, a psychiatric dis-
order, a history of diabetes with impaired renal function
(creatinine .106), a BMI of .40, or were unable or unwilling
to use PCA.
Downloaded from http://bja.oxfordjournals.org/ by guest on October 14, 2016
All subjects were screened in order to ensure eligibility
and patients were recruited at the preoperative assessment
12 weeks before surgery. At that time, the study protocol,
use of the PCA pump, and the visual analogue scale (VAS), a
010 cm scale (with endpoints labelled no pain and worst
pain possible pain), were explained. Baseline physical function
measures and psychosocial questionnaires were completed at
the preoperative assessment or on the morning of surgery
before the procedure.
Drug preparation, dispensing, and randomization
Gabapentin and placebo medications were encapsulated in
identically coloured gelatin capsules and packaged in identical
individual blister packs by the Sunnybrook Health Sciences
Centre Investigational Pharmacy in order to maintain double-
blinded conditions. The placebo capsules contained a mixture
of 50% cellulose and 50% lactose monohydrate. A computer-
generated randomization schedule was used to assign patients
at random, in blocks of six, to one of the two treatment groups.
The schedule was created by the hospital investigational phar-
macy using randomization.com (http://www.jerrydallal.com/
random/randomize.htm). The investigational pharmacy was
otherwise not involved in the clinical care of the patients or in
the conduct of the trial. The randomization schedule was kept
in the pharmacy and none of the investigators had access to it.
The pharmacy dispensed the capsules according to the random-
ization schedule when the investigators informed them that a
patient had been recruited into the trial. Researchers were also
blind to drug assignment during data analysis.
Pre- and intra-operative anaesthesia care
Standard practice at the Sunnybrook Holland Orthopaedic and
Arthritic Centre is for patients to continue taking celecoxib until
surgery. On the day of surgery, all patients received celecoxib
400 mg p.o., 2 h before operation. Patients were randomly
assigned to receive either gabapentin 600 mg p.o. or placebo
p.o. at the same time they received the celecoxib. Two hours
after study medication ingestion, patients were transferred
Perioperative gabapentin and functional outcomes
to the regional anaesthesia block area where an i.v. cannula
was inserted and an infusion of i.v. lactated Ringers solution
was started at a rate of 100 ml h21. Blood pressure, ECG, and
oximetry were monitored. Midazolam 13 mg i.v. was adminis-
tered. Femoral and sciatic nerve blocks were performed using a
nerve stimulator/ultrasound technique. A motor-evoked re-
sponse at a current of ,0.5 mA was accepted as an endpoint
of stimulation. Ropivacaine 0.5%, 20 ml was deposited adja-
cent to each nerve. Spinal anaesthesia was performed in the
lateral decubitus or sitting position using 25 g Whitacre
needle. Ten milligrams of 0.5% hypobaric bupivacaine with
10 mg of fentanyl was injected. In the operating theatre, sed-
ation was provided with i.v. propofol infusion (25 100 mg kg
min21) until the end of surgery. Surgical techniques were stan-
dardized at the Holland Orthopaedic and Arthritic Centre. The
attending anaesthesiologist was not involved in the patients
evaluation after operation.
Postoperative anaesthesia care
Patients were asked to record their rest and movement-evoked
pain intensity using a 10 cm VAS commencing in the post-
anaesthetic care unit (PACU) after surgery upon being given
the i.v. PCA morphine device, and continuing every 6 h for the
next 24 h. The PCA pump was set to deliver morphine 1 mg
per demand with a 5 min lockout interval and no background
infusion. All patients were instructed to maintain their pain in-
tensity at ,4/10 on the VAS. If the VAS pain score at rest was
rated 5 cm or greater on two consecutive 4 h assessments,
the dose of i.v. PCA morphine was increased to 1.5 mg per
demand. At each time point when pain was measured, patients
were also assessed for the presence of nausea, vomiting, and
pruritus and the severity of sedation [0alert, 1mild (occa-
sionally drowsy, easy to arouse), 2moderate (frequently
drowsy, easy to arouse), 3severe (somnolent, difficult to
arouse), Snormal sleep, easy to arouse].
Upon discharge from the PACU, all patients received a stand-
ard postoperative regimen of celecoxib 200 mg q12h and a
morphine i.v. PCA device for 24 h. Patients received either gaba-
pentin 200 mg TID or placebo TID as per the preoperative ran-
domization allocation starting 8 h after the preoperative dose
and continuing for 4 days. Oxycontin 5 mg q8h was started at
08:00 on the morning after surgery to facilitate the termination
of PCA morphine at 24 h. After operation, all patients were
managed using a standardized knee care pathway that
included daily physiotherapy treatment. Patients were permit-
ted to be full weight-bearing and participated in a progressive
programme of range of motion, strengthening exercises, and
functional training.
The Departments of Anaesthesiology, Orthopaedic Surgery,
Rehabilitation Science and Psychology collaborated on the
current project, and each unit was instrumental in selecting
the various outcome measures and in ensuring appropriately
qualified people collected the data. Fully qualified and
trained physiotherapists at the Holland Orthopaedic and Arth-
ritic Centre collected the physiotherapy data and our study
coordinators (certified RNs and research coordinators) were
BJA
responsible for the collection of Pain data and follow-up ques-
tionnaire data. Blinding was maintained throughout the study
until the code was broken upon the completion of our statistical
analysis.
Questionnaires
Western Ontario and McMaster Universities Osteoarthritis
Index (WOMAC LK3.1)
The WOMAC is a 24-item Likert scale that assesses the extent to
which patients with knee, hip, or both osteoarthritis experience
pain, stiffness, and physical functional impairment.15 16 The
WOMAC comprises three subscales, namely pain (5 items),
stiffness (2 items), and physical function (17 items). For each
item, patients are asked to rate the extent to which they experi-
ence pain/stiffness/physical limitations on a scale ranging from
0 (none) to 4 (extreme). Total scores range from 0 to 20 for the
pain subscale, 0 to 8 for the stiffness subscale, and 0 to 68 for
Downloaded from http://bja.oxfordjournals.org/ by guest on October 14, 2016
the physical function subscale with higher scores indicating
worse pain/stiffness/physical limitations. The WOMAC sub-
scales have very good internal consistency (a 0.86 0.95),25
adequate testretest reliability (intraclass correlation
coefficient0.740.95),26 27 good sensitivity and responsive-
ness to change over time,28 30 and also adequate face,15 con-
struct30 32, and criterion32 34 validity. Important
within-patient change scores have been estimated to be 4.5
for pain and 9 for physical function.35 36
Hospital Anxiety and Depression Scale
The Hospital Anxiety and Depression Scale (HADS) is a 14-item
Likert scale that assesses symptoms of depression (7-item sub-
scale) and anxiety (7-item subscale).37 The HADS is the pre-
ferred measure of anxiety and depressive symptoms for
non-psychiatric hospital patients. For each item, patients are
asked to check the answer that most closely describes how
they have been feeling in the past week on a scale from 0 to
3. Total scores for each subscale range from 0 to 21, with
higher scores indicating higher levels of anxiety or depression.
The internal consistency of the depression (a 0.670.90) and
anxiety (a 0.680.93) subscales of the HADS is adequate.38
The HADS also has good discriminant and convergent validity
and also specificity and sensitivity in detecting clinical signifi-
cant levels of depression and anxiety.38 When administered
on POD4, patients were asked to report how they have been
feeling while in-hospital.
Functional outcomes
Knee range of motion, the timed get up and go test, the stair
test, and the 6MWT were used to evaluate functional perform-
ance.
Knee range of motion
Active assisted knee flexion and active knee extension were
recorded. Immediately after the measurement of knee
flexion, patients were asked to rate the intensity of their pain.
Knee range motion has been shown to have good reliability
857
BJA
among patients with knee osteoarthritis.22 Range of motion
was measured pre-surgery and daily on PODs 1, 2, 3, and 4.
Timed get up and go test
Patients were asked to rise from a standard arm chair, walk at a
safe and comfortable pace for 3 m, and then return to a sitting
position in the chair.17 18 Time to complete the test in seconds,
use of an armrest to get up and sit down, and pain intensity at
the site of surgery immediately after the test were recorded.
This test was administered pre-surgery, on POD4 (for those
who were able to perform it), and 6 weeks and 3 months post-
surgery. An important within-patient change has been esti-
mated to be 2.5 s.22 39
Stair test
Patients were asked to ascend and descend one flight of nine
stairs in their usual manner, at a safe and comfortable
pace.19 40 Test completion, length of time to complete the
test, aids used (crutches, cane, railing), ascending and des-
cending pattern, and also pain intensity at the site of surgery
were recorded. This test was administered pre-surgery, on
POD4 (for those who were able to perform it), and 6 weeks
and 3 months post-surgery. An important within-patient
change has been estimated to be 5.5 s.22 39
Six minute walk test
Patients were asked to walk as far as possible during 6 min, with
the option to stop and rest as needed during the test.41 43
Standardized encouragement was offered to all patients
every 60 s as research has shown that encouragement
improves performance.44 Total distance walked, test comple-
tion, and pain intensity at the site of surgery were recorded.
This test was administered pre-surgery and at 6 weeks and
3 months post-surgery. An important within-patient change
has been estimated to be 62 m.22 39
Sample size estimate
There is a paucity of information available concerning the mag-
nitude of a clinically important between-group difference
for performance measures. For this reason, our sample size
estimate was based on the WOMAC physical function scores.
Goldsmith and colleagues have suggested that for self-report
measures an important between-group difference is less than
an important within-patient change.45 Applying this principle
to the WOMAC physical function scale, we specified an import-
ant between-group difference to be 5-points at 4 days and 6
weeks and 4-points at 3 months post-arthroplasty. The power
was set at 0.80 and the overall Type I error probability at 0.05
corrected to 0.0167 owing to comparisons at three time points
(POD4, 6 weeks, and 3 months). Applying these assumptions
yielded an approximate sample size of 72 subjects per group.
Allowing for an incomplete data percentage of 20%, 92 subjects
per group or 184 subjects in total were required.
Statistical analysis
Given a repeated measures study design, the likelihood of
missing data, correlated errors within individuals, and
858
Clarke et al.
heterogeneity among occasion variances (i.e. over time), we
applied Generalized Estimating Equations to test for differences
in the TUG test, the stair test, and the 6MWT between the
Groups G and P. Dependent variables were the outcome mea-
sures assessed at multiple time points (i.e. TUG test, stair test,
and 6MWT). The independent variable was treatment group
(gabapentin or placebo) and the covariates were gender, age,
and the pre-surgery values for the dependent variable of interest.
We applied an autoregressive correlation structure for cumula-
tive morphine consumption and an unstructured correlation
structure for all other dependent variables. An effect was consid-
ered statistically significant at P0.05 and we performed
intention-to-treat analyses. We did not impute values when
data were missing. All analyses were conducted using STATA
version 13.0 (STATA Corp., College Station, TX, USA).
Results
Recruitment and retention of patients
Downloaded from http://bja.oxfordjournals.org/ by guest on October 14, 2016
Patients were recruited between November 2007 and March
2011. The CONSORT46 flow chart outlining the recruitment and
retention of study patients is shown in Fig. 1. Overall, 650 patients
were approached for participation, 438 patients declined partici-
pation and 212 patients consented to participate in the study, 33
of whom were excluded or withdrew before randomization. Of
the 179 patients randomly assigned to receive gabapentin
(Group G) or placebo (Group P), 165 patients remained in the
study 24 h after operation, 77 of 84 patients (92%) in Group P,
and 88 of 95 (93%) patients remained in Group G. The reasons
for withdrawal/dropout in Group P at 24 h were as follows:
three patients had their surgery cancelled after the study medi-
cations were given, two patients were removed byattending phy-
sicians after being given general anaesthesia, and two patients
asked to be removed on POD 1 because they were unhappy
with their pain control. The reasons for withdrawal/dropout in
Group G at 24 h were as follows: four patients had their surgery
cancelled after the study medications were given and three
patients asked to be removed early on POD1 because they
were unhappy with their pain control. On POD4, 150 patients
were eligible to complete the physiotherapy measures. Six
weeks after surgery data were collected from 76 patients in the
placebo group (one patient was lost to follow-up) and 81 patients
in the gabapentin group (seven patients were lost to follow-up).
At 3 months after surgery, data were collected from 76 patients
in the placebo group and 79 patients in the gabapentin group
(two patients were lost to follow-up).
Baseline characteristics and clinical variables
The two groups were comparable with respect to age, sex, BMI,
ASA status, duration of surgery, and all self-report scales and
physical functional assessment measures (Table 1).
Performance and patient-report measures/primary
outcomes
Performance outcomes on the TUG test, the stair test, the
6MWT, and WOMAC function scores are presented in Table 2 for
the POD4, 6 week, and 3 month assessments. Significant
Perioperative gabapentin and functional outcomes BJA

Approached
(n=650)
438 patients refused
participation
Consented
n=212
33 excluded: upon receiving blood
work ineligible (elevated creatinine)
or withdrew consent over the phone
prior to surgery
Randomized
(n=179)

Group P Group G
Excluded:
Placebo Gabapentin Excluded:
3 surgeries cancelled
(n=84) 600 mg (n=95) 4 surgeries cancelled
2 excluded given GA
3 withdrew on POD1
2 withdrew on POD1
unhappy with
unhappy with pain
pain control
control 24 h: 24 h:

Downloaded from http://bja.oxfordjournals.org/ by guest on October 14, 2016

n=77 n=88

POD4: POD4:
n=70 n=80

1 lost to follow up 7 lost to follow up

6 week follow 6 week follow

up: n=76 up:n=81

2 lost to follow up

3 month 3 month
follow up: follow up:
n=76 n=79

Fig 1 Patient enrolment and study flow. GA, general anaesthesia.

differences were not found on any performance-based measure
at these time points nor was there a difference in patient-
reported function across the time points. Moreover, the confi-
dence intervals (CIs) on the WOMAC physical function between-
group difference at 6 weeks and 3 months did not include our
declared clinically important difference value of 5 points.
Table 3 presents the pain scores associated with each func-
tional measure and also the WOMAC pain scores for the POD4,
6 week, and 3 month assessments. Significant differences were
not found on any pain measure associated with the three
performance-based measures at any point in the study.
In-hospital knee range of motion
Table 4 shows the knee flexion (range of motion) and asso-
ciated pain scores, occasion-specific sample sizes, unadjusted
group means and standard errors, and adjusted between-
group differences with accompanying 95% CI and P-values.
The gabapentin group had significantly better active assisted
knee range of motion during the first 3 days of hospital rehabili-
tation compared with controls; however, this difference was
not maintained at subsequent assessments. Pain scores were
not different between the two groups during the physiotherapy
sessions.

Secondary outcomes
Opioid consumption
Cumulative morphine consumption was significantly lower in
the gabapentin group compared with the placebo group at
18 and at 24 h after TKA (P,0.0125) (Fig. 2). Pain scores did
not differ significantly at rest or with movement between the
two groups over the first 24 h.
Adverse effects
On POD1, the placebo group had a higher incidence of
nausea (Group P 34% vs Group 30% G) (P0.013) and pruritus
(Group P 31% vs Group G 4%) (P0.007). On POD3, the
placebo group also reported more dizziness (Group P 15% vs
Group G 2%) (P0.025) compared with the gabapentin
group. Significant differences were not found at any point

859
BJA Clarke et al.

during the hospital stay with respect to sedation severity Hospital Anxiety and Depression Scores on POD4, 6 weeks,
or the incidence of vomiting. No other adverse events were and 3 months
documented. HADS anxiety scores did not differ between groups on POD4
[G: 5.9 (2.9), P: 5.2 (3.3)] and 6 weeks [G: 5.2 (3.5), P: 4.3 (3.5)]
or 3 months [G: 3.9 (3.2), P: 3.5 (2.8)] after TKA. Similarly,
Table 1 Patient, baseline characteristics and duration of surgery HADS depression scores did not differ between groups on
POD4 [G: 4.6 (3.0), P: 4.8 (3.1)] and 6 weeks [G: 4.2 (2.9), P: 3.8
Variable Group P-value
(3.0)] and 3 months [G: 5.2 (2.2), P: 5.6 (2.4)] after surgery
Placebo Gabapentin
(P.0.05).
(n584) (n595)
Age (yr) 62.8 (7.4) 62.9 (6.3) 0.92
Sex (male/female) 47/37 42/53 0.11
Discussion
Body mass index (kg m 22
) 31.1 (5.6) 32.3 (5.1) 0.13 The pain-relieving and opioid-sparing effects of gabapentin
Duration of surgery (min) 71.3 (22.0) 71.8 (19.7) 0.87 have been studied more frequently among patients receiving
ASA class n.s. only general anaesthesia than in those receiving regional an-
I 8 8 aesthesia. The present study was designed with the primary
II 59 68 aim of examining whether a 4 day perioperative regimen of
III 17 19 gabapentin within the context of spinal anaesthesia, periph-
eral nerve blocks (femoral and sciatic), perioperative adminis-

Downloaded from http://bja.oxfordjournals.org/ by guest on October 14, 2016

Knee flexion range (8) 113.6 (14.1) 112.3 (17.1) 0.58
Stair test (s) 20.8 (9.5) 22.4 (11.3) 0.31 tration of NSAIDs (celecoxib), and i.v. PCA morphine for 24 h
Stair pain (0 10) 4.4 (2.3) 4.6 (2.5) 0.58 would improve knee function and functional recovery post-
TUG (s) 12.5 (4.5) 12.2 (4.7) 0.66 discharge. The 600 mg preoperative dose of gabapentin was
TUG pain (0 10) 3.4 (2.3) 3.2 (2.3) 0.56 chosen, given a study that followed patients undergoing
6MWT (m) 356 (131) 358 (136) 0.92 lumbar discectomy suggested that the optimal preoperative
6MWT pain (0 10) 5.5 (2.3) 5.0 (2.3) 0.14 dose of gabapentin for optimal acute postoperative pain
WOMAC pain (020) 9.8 (2.9) 9.8 (3.3) .0.99 relief was 600 mg; at higher doses (900 and 1200 mg), an anal-
WOMAC stiffness (0 8) 4.3 (1.6) 4.7 (1.6) 0.10 gesic ceiling effect was observed, in which patients exhibited
WOMAC physical function 32.6 (9.2) 34.4 (12.3) 0.27 more side-effects without an additional reduction in acute
(0 68) pain.47 At the time of commencement of this trial, no study
HADSAnxiety (0 21) 6.0 (3.8) 6.0 (3.1) .0.99 had been published using gabapentin in an awake population
HADSDepression (0 21) 4.5 (2.5) 4.2 (2.6) 0.43 with a robust multimodal regimen which included preoperative
spinal anaesthesia, NSAIDs (celecoxib), and peripheral femoral

Table 2 Performance and patient-reported physical function measures. *Negative difference favours gabapentin group; positive difference
favours gabapentin group. WOMAC scores: Physical Function Subscale. P-value reflects ANCOVA

Group Adjusted difference

Placebo Gabapentin (95% CI) P-value
Unadjusted Unadjusted
n Mean (SE) n Mean (SE)
TUG* (s)
POD4 66 42.6 (3.6) 71 40.1 (3.4) 23.0 (28.4, 2.5) 0.286
6 weeks 77 13.5 (0.9) 81 13.8 (0.7) 20.1 (25.2, 4.9) 0.956
3 months 71 10.6 (0.7) 79 10.4 (0.5) 20.5 (25.7, 4.9) 0.850
Stair test* (s)
POD4 57 56.9 (4.4) 62 57.5 (3.3) 22.3 (28.4, 3.7) 0.451
6 weeks 72 22.7 (1.5) 79 24.2 (1.5) 0.6 (24.8, 6.0) 0.829
3 months 69 15.8 (1.5) 77 18.4 (1.5) 0.2 (25.2, 5.7) 0.941
Six minute walk (m)
6 weeks 75 333 (16) 76 363 (15) 27.5 (25.9, 60.9) 0.107
3 months 69 436 (15) 78 444(12) 9.4 (224.6, 43.4) 0.588
WOMAC function* (0 68)
POD4 44 35.3 (1.4) 51 33.8 (1.3) 22.2 (26.3, 2.0) 0.302
6 weeks 72 19.9 (1.3) 81 23.8 (1.4) 3.1 (20.2, 6.5) 0.065
3 months 76 13.6 (1.2) 79 15.2 (1.3) 1.0 (22.2, 4.3) 0.540

860
Perioperative gabapentin and functional outcomes BJA

Table 3 Performance-related numeric pain rating scale (0 10) scores. *Negative difference favours gabapentin group; positive difference favours
gabapentin group. P-value reflects ANCOVA

Group Adjusted difference

Placebo Gabapentin (95% CI) P-value
Unadjusted Unadjusted
n Mean (SE) n Mean (SE)
TUGpain*
POD4 65 4.2 (0.3) 72 4.3 (0.3) 0.1 (20.5, 0.7) 0.836
6 weeks 74 1.7 (0.2) 81 2.0 (0.2) 0.2 (20.3, 0.8) 0.436
3 months 72 0.7 (0.2) 78 0.8 (0.2) 0.1 (20.5, 0.7) 0.690
Stair testpain*
POD4 61 4.0 (0.3) 63 4.3 (0.3) 0.2 (20.4, 0.9) 0.488
6 weeks 74 2.2 (0.2) 77 2.8 (0.2) 0.4 (20.2, 1.0) 0.154
3 months 69 1.2 (0.2) 76 1.6 (0.2) 0.3 (20.3, 0.9) 0.299
Six minute walkpain
6 weeks 72 2.9 (0.2) 73 2.9 (0.2) 0.2 (20.4, 0.8) 0.475

Downloaded from http://bja.oxfordjournals.org/ by guest on October 14, 2016

3 months 76 1.2 (0.2) 75 1.4 (0.2) 0.2 (20.4, 0.9) 0.435
WOMACpain* (0 20)
POD4 69 8.5 (0.4) 80 8.3 (0.3) 20.5 (21.4, 0.5) 0.342
6 weeks 76 6.2 (0.3) 81 7.3 (0.4) 1.0 (0.1, 1.9) 0.039
3 months 76 3.5 (0.5) 79 4.2 (0.4) 0.7 (20.2, 1.7) 0.133

60 P=0.459 P=0.047 P=0.004 P=0.001

50
Morphine consumption (mg)

40

30

20

Placebo
10
Gabapentin

0
6 12 18 24
Time after surgery (h)

Fig 2 Twenty-four-hour cumulative morphine consumption [mean (95% CI)].

and sciatic nerve blocks and data were available which demon-
strated the effectiveness of this regimen for the reduction of
opioid consumption and improvement in early knee function
after TKA.9 The present study did not find significant differ-
ences between gabapentin and placebo groups in the
primary measures of function (both physical and patient
reported) on POD4 and 6 weeks or 3 months after TKA
(Tables 2 and 3). Furthermore, pain scores at 6 weeks and 3
months did not differ between groups with respect to each
performance-based measure. Consistent with previous
literature,9 12 this study demonstrated a 24 h reduction in
opioid consumption and an overall improvement in active
assisted knee flexion to 4 days after operation while patients
were receiving gabapentin (Fig. 2 and Table 4).
The possibility that gabapentin may have preventive peri-
operative analgesic effects (as defined by demonstrating
pain reduction, improved function, or both when compared
with controls .5.5 biological half-life after the medication)48
has been reported in previous studies.49 50 A published
meta-analysis found that 50% of the published trials to date

861
BJA Clarke et al.

Table 4 Knee flexion range of motion and pain scores. *Negative difference favours gabapentin group; positive difference favours gabapentin
group. P-value reflects ANCOVA

Group Adjusted difference

Placebo Gabapentin (95% CI) P-value
Unadjusted Unadjusted
n Mean (SE) n Mean (SE)
Knee flexion range (8)
POD1 76 55.1 (2.2) 83 60.8 (1.8) 5.9 (0.9, 11.0) 0.020
POD2 74 62.8 (1.9) 85 69.1 (1.4) 7.9 (2.9, 13.0) 0.002
POD3 77 75.0 (1.6) 84 79.5 (1.5) 5.9 (0.9, 10.9) 0.021
POD4 70 80.9 (1.3) 80 83.8 (1.3) 4.1 (21.1, 9.3) 0.123
6 weeks 74 102.5 (2.1) 81 105.1 (1.4) 0.9 (24.2, 6.0) 0.724
3 months 74 115.2 (1.6) 79 116.2 (1.3) 0 (25.1, 5.1) 0.997
Knee flexion pain* (0 10)
POD1 76 6.4 (0.2) 83 5.8 (0.3) 20.4 (21.2, 0.4) 0.336
POD2 74 5.6 (0.3) 85 6.0 (0.2) 0.8 (0, 1.6) 0.050

Downloaded from http://bja.oxfordjournals.org/ by guest on October 14, 2016

POD3 77 5.7 (0.3) 84 5.5 (0.3) 0.2 (20.6, 1.0) 0.665
POD4 70 5.5 (0.3) 80 5.6 (0.3) 0.5 (20.3, 1.3) 0.255
6 weeks 70 2.3 (0.2) 75 2.6 (0.2) 0.3 (20.6, 1.2) 0.493
3 months 70 1.0 (0.2) 74 1.3 (0.2) 0.6 (20.2, 1.5) 0.139

which examined the use of gabapentin for the reduction of
chronic post-surgical pain were positive, but clearly more
evidence is needed to determine whether this class of medica-
tions has such reported preventive effects.24 Furthermore,
others have suggested that early reductions in acute post-
surgical pain could lead to improvements in early mobilization
and this head start in activity might improve the long-term
trajectory of treated patients.51 52 We designed this trial to de-
termine whether gabapentin would demonstrate preventive
analgesic effects and included time points well beyond the
physiologically active duration of the medication (i.e. 6 weeks
and 3 months after surgery). The present results do not
support the hypothesis of superior functional performance
long-term in the gabapentin vs the placebo group.
The results of the present study confirm open label evi-
dence9 that perioperative gabapentin when added to cele-
coxib for 4 days after operation reduces opioid consumption
and leads to an improvement in active assisted knee flexion
while patients are in-hospital. Consistent with previous
results,9 there was no difference in opioid consumption
between the two groups in the first 12 h after surgery, which
demonstrates the effectiveness of the femoral and sciatic
nerve blocks in the first 812 h after total knee replacement.
Active assisted knee flexion in the gabapentin group was
6.18 greater than the placebo group during the first 4 days
after surgery. The significantly greater incidence of pruritus
and nausea found on POD1 in the placebo group is associated
with increased morphine use by this group and is consistent
with previous studies.53
Patients self-report measures of anxiety and depressive
symptoms throughout the trial were unaffected by gabapentin
administration. Studies have shown that a high level of
preoperative anxiety is a risk factor for acute post-surgical
pain54 and is associated with the development of chronic post-
surgical pain.55 In a select population of highly anxious females
(i.e. baseline anxiety .5/10) undergoing major surgery, 1200
mg of gabapentin significantly reduced preoperative anxiety
and pain catastrophizing before major surgery.56 The dosing
regimen of the present study (i.e. 600 mg) and the low baseline
levels of anxiety symptoms may explain the absence of an
anxiolytic effect of gabapentin.
This is the first adequately powered study to compare the
efficacy of gabapentin vs placebo for recovery of physical
function over three periods (POD4, 6 weeks, and 3 months)
after TKA. Although negative for such outcomes, a reduction
in cumulative opioid consumption and superior in-hospital
range of motion recovery were evident. These latter findings
are consistent with the bulk of the published literature in
patients undergoing surgeries associated with moderate/
severe post-surgical pain. Other populations and regimens
examining the use of perioperative gabapentin are needed.
The present results are specific to our multimodal analgesic
regimen and cannot be generalized to the same surgery with
alternative perioperative pain regimens or other surgeries per-
formed under general anaesthesia. Furthermore, given the low
rate of consent (32%) of patients that were approached and
the 14% of patients that were lost to follow-up, the generaliz-
ability of this study may be limited.
In conclusion, a single preoperative dose of gabapentin 600
mg, co-administered with celecoxib, and followed by 4 days of
gabapentin 200 mg TID, decreased immediate postoperative
analgesic requirements and resulted in improved range of
motion while patients were taking active medication and
undergoing in-hospital rehabilitation. Gabapentin provided

862
Perioperative gabapentin and functional outcomes
no improvement in pain or physical function on POD4, 6 weeks,
or 3 months after surgery. The effects on functional outcomes
based on the continued use of gabapentin into the post-
discharge period need to be assessed in future work. Further-
more, future studies should continue to explore potential
patient populations (i.e. patients with obstructive sleep
apnoea, chronic pain patients, opioid dependent, etc.) that
may benefit from the early opioid-sparing benefits and the
early improvement of physical function associated with the ad-
ministration of perioperative gabapentin.
Authors contributions
H.C. is the primary author of this manuscript and responsible for
designing the study and reviewing the analysis of the data, and
has approved the final manuscript. J.K. is a co-author of the
manuscript responsible for helping to design the study,
writing some of the manuscript, and approving the final manu-
script. C.McC. is a co-author of the manuscript responsible for
overseeing the study at the Holland Centre, writing some of
the manuscript, and approving the final manuscript. P.S. is
the biostatistician responsible for the analysis of the data
and has approved the final manuscript. D.K. contributed to
the writing of the manuscript, helped to develop the initial
protocol, and has approved the final manuscript. G.P. contribu-
ted to the manuscript, helped with data analyses, and
approved the final manuscript. I.T.A. is a co-author of the
manuscript, responsible for overseeing the study at the
Holland Centre, writing some of the manuscript, and approving
the final manuscript. J.G. is an orthopaedic surgeon, contribu-
ted to the initial design of the study, and contributed to and
approved the final manuscript. J.K. is the senior author and
was involved in the design of the study, analysis of the
results, and writing of the manuscript. He has approved the
final manuscript.
Acknowledgements
We thank research coordinators Beth Goudie and Ashley Pope,
the Nursing Staff, Pharmacy, and the Physiotherapy teams at
the Holland Orthopaedic and Arthritic Centre, Noga Cohen,
and the Pain Research team at the Toronto General Hospital
for their hard work and support.
Declaration of interest
None declared.
Funding
H.C. and C.McC. are supported by Merit Awards from the Depart-
ment of Anaesthesia at the University of Toronto. H.C. is
supported by the STAGE Training Program in Genetic Epidemi-
ology from the Canadian Institutes of Health Research. J.K. is
supported by a Canada Research Chair in Health Psychology
at York University. This study was made possible through
grants by the Canadian Anesthesiology Society, the Sunny-
brook Health Sciences Centre Practice Based Research Fund,
and a grant from the Orthopedic Foundation of Canada.
BJA
References
1 Singh JA, Vessely MB, Harmsen WS, et al. A population-based study
of trends in the use of total hip and total knee arthroplasty, 1969
2008. Mayo Clin Proc 2010; 85: 898 904
2 Lawrence RC, Felson DT, Helmick CG, et al. Estimates of the preva-
lence of arthritis and other rheumatic conditions in the United
States. Part II. Arthritis Rheum 2008; 58: 26 35
3 Mahomed NN, Barrett J, Katz JN, Baron JA, Wright J, Losina E. Epi-
demiology of total knee replacement in the United States Medicare
population. J Bone Joint Surg 2005; 87: 1222 8
4 Ho KY, Gan TJ, Habib AS. Gabapentin and postoperative paina sys-
tematic review of randomized controlled trials. Pain 2006; 126:
91 101
5 Peng PW, Wijeysundera DN, Li CC. Use of gabapentin for periopera-
tive pain controla meta-analysis. Pain Res Manag 2007; 12:
85 92
6 Seib RK, Paul JE. Preoperative gabapentin for postoperative anal-
gesia: a meta-analysis. Can J Anaesth 2006; 53: 4619
7 Backes JR, Bentley JC, Politi JR, Chambers BT. Dexamethasone
reduces length of hospitalization and improves postoperative
Downloaded from http://bja.oxfordjournals.org/ by guest on October 14, 2016
pain and nausea after total joint arthroplasty: a prospective, rando-
mized controlled trial. J Arthroplasty 2013; 28 (8 Suppl): 11 7
8 Horlocker TT. Pain management in total joint arthroplasty: a histor-
ical review. Orthopedics 2010; 33(9 Suppl.): 14 9
9 Clarke H, Pereira S, Kennedy D, et al. Gabapentin decreases mor-
phine consumption and improves functional recovery following
total knee arthroplasty. Pain Res Manag 2009; 14: 217 22
10 Paul JE, Nantha-Aree M, Buckley N, et al. Gabapentin does not
improve multimodal analgesia outcomes for total knee arthroplasty:
a randomized controlled trial. Can J Anaesth 2013; 60: 42331
11 Gilron I, Orr E, Tu D, ONeill JP, Zamora JE, Bell AC. A placebo-
controlled randomized clinical trial of perioperative administration
of gabapentin, rofecoxib and their combination for spontaneous
and movement-evoked pain after abdominal hysterectomy. Pain
2005; 113: 191200
12 Menigaux C, Adam F, Guignard B, Sessler DI, Chauvin M. Preopera-
tive gabapentin decreases anxiety and improves early functional
recovery from knee surgery. Anesth Analg 2005; 100: 13949
13 Turan A, White PF, Karamanlioglu B, et al. Gabapentin: an alterna-
tive to the cyclooxygenase-2 inhibitors for perioperative pain man-
agement. Anesth Analg 2006; 102: 175 81
14 Stratford PW, Kennedy DM. Performance measures were necessary
to obtain a complete picture of osteoarthritic patients. J Clin Epide-
miol 2006; 59: 1607
15 Bellamy N, Buchanan WW. A preliminary evaluation of the dimen-
sionality and clinical importance of pain and disability in osteoarth-
ritis of the hip and knee. Clin Rheumatol 1986; 5: 23141
16 Bellamy N. WOMAC Osteoarthritis Index User Guide. Version VII. Bris-
bane, Australia, 2005
17 Freter SH, Fruchter N. Relationship between timed up and go and
gait time in an elderly orthopaedic rehabilitation population. Clin
Rehabil 2000; 14: 96 101
18 Ouellet D, Moffet H. Locomotor deficits before and two months after
knee arthroplasty. Arthritis Rheum 2002; 47: 48493
19 Kennedy D, Stratford PW, Pagura SM, Walsh M, Woodhouse LJ. Com-
parison of gender and group differences in self-report and physical
performance measures in total hip and knee arthroplasty candi-
dates. J Arthroplasty 2002; 17: 70 7
20 Walsh M, Kennedy D, Stratford PW, Woodhouse LJ. Perioperative
functional performance of women and men following total knee
arthroplasty. Physiother Can 2001; 53: 92 100
863
BJA
21 Kreibich DN, Vaz M, Bourne RB, et al. What is the best way of asses-
sing outcome after total knee replacement? Clin Orthop Relat Res
1996; 2215
22 Kennedy DM, Stratford PW, Wessel J, Gollish JD, Penney D. Assessing
stability and change of four performance measures: a longitudinal
study evaluating outcome following total hip and knee arthro-
plasty. BMC Musculoskelet Disord 2005; 6: 3
23 Paul JE, Arya A, Hurlburt L, et al. Femoral nerve block improves an-
algesia outcomes after total knee arthroplasty: a meta-analysis of
randomized controlled trials. Anesthesiology 2010; 113: 1144 62
24 Clarke H, Bonin RP, Orser BA, Englesakis M, Wijeysundera DN, Katz J.
The prevention of chronic postsurgical pain using gabapentin and
pregabalin: a combined systematic review and meta-analysis.
Anesth Analg 2012; 115: 42842
25 Bellamy N, Buchanan WW, Goldsmith CH, Campbell J, Stitt LW. Val-
idation study of WOMAC: a health status instrument for measuring
clinically important patient relevant outcomes to antirheumatic
drug therapy in patients with osteoarthritis of the hip or knee. J
Rheumatol 1988; 15: 1833 40
26 Roos EM, Klassbo M, Lohmander LS. WOMAC osteoarthritis index.
Reliability, validity, and responsiveness in patients with arthrosco-
pically assessed osteoarthritis. Western Ontario and MacMaster
Universities. Scand J Rheumatol 1999; 28: 2105
27 Dunbar MJ, Robertsson O, Ryd L, Lidgren L. Appropriate question-
naires for knee arthroplasty. Results of a survey of 3600 patients
from The Swedish Knee Arthroplasty Registry. J Bone Joint Surg Br
2001; 83: 33944
28 McConnell S, Kolopack P, Davis AM. The Western Ontario and
McMaster Universities Osteoarthritis Index (WOMAC): a review of
its utility and measurement properties. Arthritis Rheum 2001; 45:
453 61
29 Wright JG, Young NL. A comparison of different indices of respon-
siveness. J Clin Epidemiol 1997; 50: 23946
30 Brazier JE, Harper R, Munro J, Walters SJ, Snaith ML. Generic and
condition-specific outcome measures for people with osteoarth-
ritis of the knee. Rheumatology (Oxford). 1999; 38: 870 7
31 Roos EM, Roos HP, Lohmander LS. WOMAC Osteoarthritis Index
additional dimensions for use in subjects with post-traumatic
osteoarthritis of the knee. Western Ontario and MacMaster Univer-
sities. Osteoarthritis Cartilage 1999; 7: 21621
32 Bombardier C, Melfi CA, Paul J, et al. Comparison of a generic and
a disease-specific measure of pain and physical function after
knee replacement surgery. Med Care 1995; 33(4 Suppl.): AS131 44
33 Bullens PH, van Loon CJ, de Waal Malefijt MC, Laan RF, Veth RP.
Patient satisfaction after total knee arthroplasty: a comparison
between subjective and objective outcome assessments.
J Arthroplasty 2001; 16: 7407
34 Robertsson O, Dunbar MJ. Patient satisfaction compared with
general health and disease-specific questionnaires in knee arthro-
plasty patients. J Arthroplasty 2001; 16: 476 82
35 Halket A, Stratford PW, Kennedy DM, Woodhouse LJ, Spadoni G.
Measurement properties of performance-specific pain ratings of
patients awaiting total joint arthroplasty as a consequence of
osteoarthritis. Physiother Can 2008; 60: 25563
36 Pua YH, Cowan SM, Wrigley TV, Bennell KL. The lower extremity func-
tional scale could be an alternative to the Western Ontario and
McMaster Universities Osteoarthritis Index physical function
scale. J Clin Epidemiol 2009; 62: 110311
37 Zigmond AS, Snaith RP. The hospital anxiety and depression scale.
Acta Psychiatr Scand 1983; 67: 361 70
38 Bjelland I, Dahl AA, Haug TT, Neckelmann D. The validity of the Hos-
pital Anxiety and Depression Scale. An updated literature review.
J Psychosomatic Res 2002; 52: 6977
864
Clarke et al.
39 Kennedy DM, Stratford PW, Hanna SE, Wessel J, Gollish JD. Modeling
early recovery of physical function following hip and knee arthro-
plasty. BMC Musculoskelet Disord 2006; 7: 100
40 Walsh M, Woodhouse LJ, Thomas SG, Finch E. Physical impairments
and functional limitations: a comparison of individuals 1 year after
total knee arthroplasty with control subjects. Phys Ther 1998; 78:
248 58
41 Boardman DL, Dorey F, Thomas BJ, Lieberman JR. The accuracy of
assessing total hip arthroplasty outcomes: a prospective correl-
ation study of walking ability and 2 validated measurement
devices. J Arthroplasty 2000; 15: 200 4
42 Laupacis A, Bourne R, Rorabeck C, et al. The effect of elective total
hip replacement on health-related quality of life. J Bone Joint
Surg 1993; 75: 161926
43 Parent E, Moffet H. Comparative responsiveness of locomotor tests
and questionnaires used to follow early recovery after total knee
arthroplasty. Arch Phys Med Rehabil 2002; 83: 70 80
44 Guyatt GH, Pugsley SO, Sullivan MJ, et al. Effect of encouragement
on walking test performance. Thorax 1984; 39: 818 22
45 Goldsmith CH, Boers M, Bombardier C, Tugwell P. Criteria for clinical-
Downloaded from http://bja.oxfordjournals.org/ by guest on October 14, 2016
ly important changes in outcomes: development, scoring and
evaluation of rheumatoid arthritis patient and trial profiles.
OMERACT Committee. J Rheumatol 1993; 20: 561 5
46 Moher D, Schulz KF, Altman D. The CONSORT statement: revised
recommendations for improving the quality of reports of parallel-
group randomized trials. J Am Med Assoc 2001; 285: 198791
47 Pandey CK, Navkar DV, Giri PJ, et al. Evaluation of the optimal
preemptive dose of gabapentin for postoperative pain relief
after lumbar diskectomy: a randomized, double-blind, placebo-
controlled study. J Neurosurg Anesthesiol 2005; 17: 65 8
48 Katz J, Clarke H, Seltzer Z. Review article: preventive analgesia: quo
vadimus? Anesth Analg 2011; 113: 1242 53
49 Brogly N, Wattier JM, Andrieu G, et al. Gabapentin attenuates late
but not early postoperative pain after thyroidectomy with superfi-
cial cervical plexus block. Anesth Analg 2008; 107: 17205
50 Sen H, Sizlan A, Yanarates O, et al. The effects of gabapentin on
acute and chronic pain after inguinal herniorrhaphy. Eur J Anaes-
thesiol 2009; 26: 772 6
51 Katz J, Cohen L. Preventive analgesia is associated with reduced
pain disability 3 weeks but not 6 months after major gynecologic
surgery by laparotomy. Anesthesiology 2004; 101: 169 74
52 Gottschalk A, Smith DS, Jobes DR, et al. Preemptive epidural
analgesia and recovery from radical prostatectomy: a randomized
controlled trial. J Am Med Assoc 1998; 279: 107682
53 Sheen MJ, Ho ST, Lee CH, Tsung YC, Chang FL. Preoperative gabapen-
tin prevents intrathecal morphine-induced pruritus after orthoped-
ic surgery. Anesth Analg 2008; 106: 1868 72
54 Pinto PR, McIntyre T, Ferrero R, Almeida A, Araujo-Soares V.
Predictors of acute postsurgical pain and anxiety following
primary total hip and knee arthroplasty. J Pain 2013; 14: 50215
55 Theunissen M, Peters ML, Bruce J, Gramke HF, Marcus MA. Preopera-
tive anxiety and catastrophizing: a systematic review and
meta-analysis of the association with chronic postsurgical pain.
Clin J Pain 2012; 28: 819 41
56 Clarke H, Kirkham KR, Orser BA, et al. Gabapentin reduces preopera-
tive anxiety and pain catastrophizing in highly anxious patients
prior to major surgery: a blinded randomized placebo-controlled
trial. Can J Anaesth 2013; 60: 43243
Handling editor: A. R. Absalom